OAI/Analysis/KadisonKastler/MatrixNearCommutant.lean

GitHub source file

7 declarations, 0 contain sorry, 0 declared as axioms

Commutant estimates from individual matrix approximants

A matrix approximant from N commutes entrywise with a contraction in N′. The complete commutator is therefore bounded by twice the matrix error. The commutant distance formula gives distance at most that error to M′; normalization costs a further factor two for unit balls. No linear map selecting the matrix approximants is assumed.

  1. theorem OAI.KadisonKastler.MatrixNearInclusion.scaled_square line 38

    Source documentation

    Individual approximants scale with the norm of the input matrix.

    theorem scaled_square (h : MatrixNearInclusion M N γ)
        (n : ℕ) (X : CStarMatrix (Fin n) (Fin n) (H →L[ℂ] H))
        (hX : ∀ i j, X i j ∈ M) :
        ∃ Y : CStarMatrix (Fin n) (Fin n) (H →L[ℂ] H),
          (∀ i j, Y i j ∈ N) ∧ ‖X - Y‖ ≤ γ * ‖X‖ := by
    …

    Mentions: OAI.KadisonKastler.MatrixNearInclusion

  2. theorem OAI.KadisonKastler.MatrixNearInclusion.mapped_derivation_eq line 62
  3. theorem OAI.KadisonKastler.MatrixNearInclusion.commutant_complete_bound line 80

    Source documentation

    Matrix proximity bounds the full derivation of a neighboring commutant contraction, uniformly at every rectangular matrix level.

    theorem commutant_complete_bound (h : MatrixNearInclusion M N γ)
        (hγ : 0 ≤ γ) (a : H →L[ℂ] H) (ha : a ∈ N.commutant) (han : ‖a‖ ≤ 1) :
        CompletelyBoundedBy
          ((innerDerivation a).comp M.toStarSubalgebra.subtype.toLinearMap) (2 * γ) := by
    …

    Mentions: OAI.KadisonKastler.MatrixNearInclusion, OAI.KadisonKastler.CompletelyBoundedBy, OAI.KadisonKastler.innerDerivation

  4. theorem OAI.KadisonKastler.MatrixNearInclusion.commutant_infDist_le line 113

    Source documentation

    The factor one-half in the complete commutator distance formula cancels the factor two in the preceding estimate.

    theorem commutant_infDist_le (h : MatrixNearInclusion M N γ)
        (hγ : 0 ≤ γ)
        (hPublished : PublishedBlockCommutantApproximation (Algebra M) H)
        (a : H →L[ℂ] H) (ha : a ∈ unitBall N.commutant) :
        Metric.infDist a (M.commutant : Set (H →L[ℂ] H)) ≤ γ := by
    …

    Mentions: OAI.KadisonKastler.MatrixNearInclusion, OAI.KadisonKastler.PublishedBlockCommutantApproximation, OAI.KadisonKastler.Algebra, OAI.KadisonKastler.unitBall

  5. theorem OAI.KadisonKastler.MatrixNearInclusion.commutant_nearInclusion line 128

    Source documentation

    The algebra near-inclusion is contravariant and needs no normalization loss. The larger constant is needed only for proximity of unit balls.

    theorem commutant_nearInclusion (h : MatrixNearInclusion M N γ)
        (hγ : 0 ≤ γ)
        (hPublished : PublishedBlockCommutantApproximation (Algebra M) H)
        {η : ℝ} (hγη : γ < η) : NearInclusion N.commutant M.commutant η :=
    …

    Mentions: OAI.KadisonKastler.MatrixNearInclusion, OAI.KadisonKastler.PublishedBlockCommutantApproximation, OAI.KadisonKastler.Algebra, OAI.KadisonKastler.NearInclusion

  6. theorem OAI.KadisonKastler.MatrixNearInclusion.exists_commutant_unitBall_approx line 136

    Source documentation

    Normalization yields a commutant unit-ball approximant with any error strictly larger than twice the matrix gap.

    theorem exists_commutant_unitBall_approx (h : MatrixNearInclusion M N γ)
        (hγ : 0 ≤ γ)
        (hPublished : PublishedBlockCommutantApproximation (Algebra M) H)
        (a : H →L[ℂ] H) (ha : a ∈ unitBall N.commutant)
        {η : ℝ} (hγη : 2 * γ < η) :
        ∃ b ∈ unitBall M.commutant, ‖a - b‖ < η := by
    …

    Mentions: OAI.KadisonKastler.MatrixNearInclusion, OAI.KadisonKastler.PublishedBlockCommutantApproximation, OAI.KadisonKastler.Algebra, OAI.KadisonKastler.unitBall

  7. theorem OAI.KadisonKastler.MatrixNearInclusion.commutant_kkDistance_le line 149

    Source documentation

    Two-sided matrix proximity controls the ordinary commutant gap.

    theorem commutant_kkDistance_le (hMN : MatrixNearInclusion M N γ)
        (hNM : MatrixNearInclusion N M γ) (hγ : 0 ≤ γ)
        (hM : PublishedBlockCommutantApproximation (Algebra M) H)
        (hN : PublishedBlockCommutantApproximation (Algebra N) H) :
        kkDistance M.commutant N.commutant ≤ 2 * γ := by
    …

    Mentions: OAI.KadisonKastler.MatrixNearInclusion, OAI.KadisonKastler.PublishedBlockCommutantApproximation, OAI.KadisonKastler.Algebra, OAI.KadisonKastler.kkDistance